基于相干点目标的多基线D-InSAR技术与地表形变监测

Surface Deformation Monitoring with Multi-Baseline D-InSAR Based on Coherent Point Target

失相干(Decorrelation)与大气波动是影响重复轨差分干涉测量(D-InSAR)进行地表形变信息提取的主要因素。相干性降低使得干涉纹图在空间上表现为不连续,难以完成相位解缠。重复观测时大气波动引起的相位延迟在空间域上的不均一分布则降低了D-InSAR提取形变信息,特别是空间范围覆盖较大的形变场的精度。介绍了一种基于相干目标的多基线D-InSAR数据处理算法。该算法根据少量SAR数据构成多基线干涉纹图集,分别利用点目标检测算法和相干系数均值作为相干目标提取的测度;利用相位回归分析模型对干涉相位进行时间域迭代处理,从干涉相位中提取线性形变速率和DEM误差改正,通过迭代处理补偿高程误差,解算线性地表形变速率。该算法提高了D-InSAR形变监测的时间采样率,能准确获取每个观测时刻的形变累积量。以沧州地区2004—2005年的SAR数据为例,获取了该地区地表沉降线性速率及其演变状况。

Decorrelation caused by temporal changes influences phase unwrapping of differential interferogram in repeat pass Differential SAR interfetometry（D-InSAR）. Phase delay due to atmosphere disturbance degrades the accuracy of D-InSAR for small deformation monitoring. In this paper, we present a Coherent Point Targets Interferometry approach to retrieval the phase history and estimate the linear deformation of a coherent scatterer. A multi-baseline interferograms algorithm based on discrete and temporarily natural or artificial reflectors is developed. In this algorithm, a multi-reference image strategy for interferograms generation is adopted to form the interferograms stack. A hi-threshold algorithm is used for coherent point targets identification. Those pixels preserving a good coherence level and presenting a point target character are identified from the whole set of interferograms. The phase regression model is used to estimate the linear deformation rate and DEM error of the coherent point targets. The method presents high flexibility with respect to the required number of images and baseline length. The technique has been tested with ENVISAT-ASAR data for land subsidence rate derivation of Cangzhou city.